Embed Size (px)
Citation preview
PNNL-19320
Prepared for the U.S. Department of Energy under Contract DE-AC05-76RL01830
Bringing Water into an Integrated Assessment Framework R. César Izaurralde Ronald D. Sands Allison M. Thomson Hugh M. Pitcher November 2010
PNNL-19320
Bringing Water into an Integrated Assessment Framework
R César Izaurralde Ronald D. Sands
Allison M. Thomson Hugh M. Pitcher
November 2010
Prepared for
the U.S. Department of Energy
under Contract DE-AC05-76RL01830
Pacific Northwest National Laboratory
Richland, Washington 99352
iii
Summary
We developed a modeling capability to understand how water is allocated within a river basin and
examined present and future water allocations among agriculture, energy production, other human
requirements, and ecological needs.
Water is an essential natural resource needed for food and fiber production, household and industrial
uses, energy production, transportation, tourism and recreation, and the functioning of natural ecosystems.
Anthropogenic climate change and population growth are anticipated to impose unprecedented pressure
on water resources during this century. Pacific Northwest National Laboratory (PNNL) researchers have
pioneered the development of integrated assessment (IA) models for the analysis of energy and economic
systems under conditions of climate change. This Laboratory Directed Research and Development
(LDRD) effort led to the development of a modeling capability to evaluate current and future water
allocations between human requirements and ecosystem services.
The Water Prototype Model (WPM) was built in STELLA®, a computer modeling package with a
powerful interface that enables users to construct dynamic models to simulate and integrate many
processes (biological, hydrological, economics, sociological). A 150,404-km2 basin in the United States
(U.S.) Pacific Northwest region served as the platform for the development of the WPM. About 60% of
the study basin is in the state of Washington with the rest in Oregon. The Columbia River runs through
the basin for 874 km, starting at the international border with Canada and ending (for the purpose of the
simulation) at The Dalles dam. Water enters the basin through precipitation and from streamflows
originating from the Columbia River at the international border with Canada, the Spokane River, and the
Snake River. Water leaves the basin through evapotranspiration, consumptive uses (irrigation, livestock,
domestic, commercial, mining, industrial, and off-stream power generation), and streamflow through The
Dalles dam. Water also enters the Columbia River via runoff from land. The model runs on a monthly
timescale to account for the impact of seasonal variations of climate, streamflows, and water uses. Data
for the model prototype were obtained from national databases and ecosystem model results.
The WPM can be run from three sources: 1) directly from STELLA, 2) with the isee Player®, or 3)
the web version of WPM constructed with NetSim® software. When running any of these three versions,
the user is presented a screen with a series of buttons, graphs, and a table. Two of the buttons provide the
user with background and instructions on how to run the model. Currently, there are five types of
scenarios that can be manipulated alone or in combination using the Sliding Input Devices: 1) interannual
variability (e.g., El Niño), 2) climate change, 3) salmon policy, 4) future population, and 5) biodiesel
production.
Overall, the WPM captured the effects of streamflow conditions on hydropower production. Under La
Niña conditions, more hydropower is available during all months of the year, with a substantially higher
availability during spring and summer. Under El Niño conditions, hydropower would be reduced, with a
total decline of 15% from normal weather conditions over the year. A policy of flow augmentation to
facilitate the spring migration of smolts to the ocean would also reduce hydropower supply. Modeled
hydropower generation was 23% greater than the 81 TWh reported in the 1995 U.S. Geological Survey
(USGS) database. The modeling capability presented here contains the essential features to conduct basin-
scale analyses of water allocation under current and future climates. Due to its underlying data structure
iv
and conceptual foundation, the WPM should be appropriate to conduct IA modeling at national and global
scales.
v
Acknowledgments
The authors wish to thank Gerry Stokes for his vision and support; Antoinette Brenkert and Jacob
Oppenheim for data collection, processing, and modeling support; and Charity Plata and Kim Swieringa
for technical editing. The work was funded by PNNL under the Laboratory Directed Research and
Development program.
vi
Acronyms and Abbreviations
aMW average Megawatts
BMRC Australian Bureau of Meteorological Research Centre
BPA Bonneville Power Administration
ENSO El Niño Southern Oscillation
EPIC Environmental Policy Integrated Climate
GCM Global Climate Model(s)
GMT global mean temperature
ha hectares
HUA hydrologic unit area(s)
HUC Hydrological Unit Code
HUMUS Hydrologic Unit Model of the United States
IA Integrated Assessment
LDRD Laboratory Directed Research and Development
NRC National Research Council
PNNL Pacific Northwest National Laboratory
UIUC University of Illinois, Urbana Champaign
U.S. United States
USDA United States Department of Agriculture
USGS United States Geological Survey
WPM Water Prototype Model
WRIA Water Resource Inventory Areas
vii
Contents
Summary ............................................................................................................................................... iii
Acknowledgments ................................................................................................................................. v
1.0 Background ................................................................................................................................... 1
1.1 Overall Objective ................................................................................................................. 1
1.2 Project Overview .................................................................................................................. 1
2.0 Problem Description ..................................................................................................................... 2
2.1 Basin Selection ..................................................................................................................... 2
2.2 Specific Objectives ............................................................................................................... 2
3.0 Description of the Study Basin ..................................................................................................... 3
3.1 Physiographic Features of the Study Basin .......................................................................... 3
3.2 Population and Economic Activity....................................................................................... 5
3.3 Agriculture ........................................................................................................................... 5
3.4 Water Resources ................................................................................................................... 7
3.5 A Simple Water Balance of the Study Basin ....................................................................... 8
3.6 Seasonal Variations of Streamflow and Spatial Distribution of Water Use ......................... 8
3.7 Electricity Generation .......................................................................................................... 12
3.8 Major Issues—Water, Energy, and Salmon ......................................................................... 14
3.8.1 Electricity .................................................................................................................. 15
3.8.2 Irrigation .................................................................................................................... 15
3.8.3 Salmon ....................................................................................................................... 15
3.8.4 Decision problem ...................................................................................................... 16
3.9 The Future—Climate Change, Legal Issues, and Demographic Changes ........................... 16
3.9.1 Climate change .......................................................................................................... 16
3.9.2 Demographics............................................................................................................ 18
3.9.3 Water rights ............................................................................................................... 18
4.0 Description of Prototype Model ................................................................................................... 19
4.1 Description of the Water Prototype Model .......................................................................... 19
4.1.1 Objective ................................................................................................................... 19
4.1.2 Modeling platform ..................................................................................................... 19
4.1.3 Model description ...................................................................................................... 19
4.1.4 Fundamental equations .............................................................................................. 22
4.1.5 Data sources .............................................................................................................. 23
4.1.6 How to run the model ................................................................................................ 23
4.1.7 Examples and discussion of selected model outputs ................................................. 24
4.1.8 Summary and future steps ......................................................................................... 26
5.0 References .................................................................................................................................... 28
viii
Figures
Figure 1. Study Basin in Eastern Washington and Northern Oregon ................................................... 3
Figure 2. Physiographic Features of the Pacific Northwest .................................................................. 4
Figure 3. Geographic Distribution of Annual Precipitation in the Study Region and Surrounding
Areas ............................................................................................................................................. 4
Figure 4. Water Balance Calculations and Additional Consumptive-use Allocations for
Watersheds of the Columbia Basin ............................................................................................... 11
Figure 5. Surface Responses of Crop Yield (Mg ha-1
) as Affected by Increases in Global Mean
Temperatures and Atmospheric CO2 Concentrations in the Columbia Basin .............................. 17
Figure 6. Surface Responses of Grape and Apple Yields (Mg ha-1
) as Affected by Increases in
Global Mean Temperatures and Atmospheric CO2 Concentrations in the Columbia Basin ........ 18
Figure 7. Model Diagram Showing the Larger Compartments and the Stocks and Flow Occurring
Throughout.................................................................................................................................... 20
Figure 8. Screenshot of the Water Prototype Model Interface .............................................................. 21
Figure 9. Predicted and Observed Streamflows (m3) at The Dalles as Affected by ENSO
Scenarios ....................................................................................................................................... 25
Figure 10. Monthly Hydropower (kWh) as Affected by ENSO Scenarios ........................................... 25
Figure 11. Simplified Version of the Bonneville Power Administration Model to Describe Smolts
Migration to the Ocean ................................................................................................................. 27
ix
Tables
Table 1. 1997 and 2002 USDA Census Data of Irrigated Crops in the Study Basin ............................ 6
Table 2. Area Harvested and Production of Various Dryland Agricultural Commodities ................... 6
Table 3. Area Harvested and Production of Various Irrigated Agricultural Commodities ................... 7
Table 4. Area Harvested and Production of Irrigated and Non-irrigated Agricultural Commodities ... 7
Table 5. Simple Annual Water Balance of the Study Basin Based on Observed Water Flows,
Observed Precipitation, Simulated Evapotranspiration, and Estimated Water Withdrawals ....... 9
Table 6. Monthly and Annual Streamflows of the Columbia River at the John Day Dam and
Monthly and Annual Water Withdrawals from the Columbia River ............................................ 10
Table 7. Locations and Characteristics of Dams Located on the Columbia River Within the Study
Region ........................................................................................................................................... 13
1
1.0 Background
1.1 Overall Objective
The overall objective of this project is to develop a quantitative understanding of how water
is allocated within a watershed among its many uses, such as: agriculture, energy production,
transportation, recreational activities, ecosystems, industrial, and municipal.
1.2 Project Overview
Water is an essential natural resource needed for food and fiber production, household and
industrial uses, energy production, transportation, tourism and recreation, and the functioning of
natural ecosystems. In most parts of the world, water resources are already subject to great stress.
During this century, climate change, population growth, increased demand for animal protein due
to rising incomes, and the potential for large-scale biomass production for climate change
mitigation are anticipated to impose unprecedented pressure on already stressed water resources.
Our integrated assessment (IA) modeling capability to analyze the impacts of climate change on
water resources has been limited to the area of water availability and demand in agriculture in the
conterminous United States (U.S.) (Izaurralde et al. 2003; Rosenberg et al. 2003; and Thomson et
al. 2005a, b, and c). These modeling studies allowed for the calculation of the potential changes
in irrigation under a variety of climate change scenarios, while taking into consideration many
biophysical processes, such as: precipitation, air temperature, evapotranspiration, runoff,
transpiration suppression, and the CO2 fertilization effect. There is a need, however, to consider
the complexity of other water issues such as water use and re-use (e.g., energy, ecosystems,
agriculture, municipal, and industrial), flow modifications (reservoirs), and water quality.
Availability of these data would be essential for integrating water into an IA modeling
framework. Our specific objective is to expand our capability to represent the multiple functions
of water in energy, economic, and environmental systems within an IA framework.
2
2.0 Problem Description
2.1 Basin Selection
Initially, we proposed to select a river basin in the southeastern U.S. and conduct a
retrospective analysis of the development of the structure and conflicts of water resources. The
criteria for basin selection included current water usage and associated conflicts, potential for
continued increase in human demands for a variety of uses, and prospects for developing an
understanding of institutional relationships governing water. In January 2006, H.M. Pitcher, A.M.
Thomson, A.L. Brenkert, R.D. Sands, and R.C. Izaurralde from the Joint Global Change Research
Institute held a teleconference with colleagues R. Skaggs, M.J. Scott, R. Leung, and L.W. Vail
from Pacific Northwest National Laboratory (PNNL) to consider the possibility of selecting a
basin in the Pacific Northwest. The rationale for the proposed selection was based on 1) a
previous water study in the Yakima Basin led by M.J. Scott, 2) regional climate simulations by R.
Leung, 3) regional hydrological modeling studies by L.W. Vail, and 4) PNNL activities in the
energy-water nexus as reported by R. Skaggs. Based on these considerations, we decided to select
a basin in the Pacific Northwest in order to develop a model prototype to analyze the role of water
within an IA framework.
2.2 Specific Objectives
The specific objective of the Laboratory Directed Research and Development (LDRD)
project was to build a model prototype that describes the uses of water, as well as their
interactions with energy systems and environmental conditions. To develop the model, we started
with a description of the study basin in terms of its biophysical, environmental, and economic
characteristics.
3
3.0 Description of the Study Basin
The basin selected is part of the Columbia Basin, a large basin covering parts of Canada and
the U.S. with two major tributaries—the Columbia and Snake rivers. The study basin was
selected to cover parts of the states of Washington and Oregon (Figure 1). It excludes part of the
Columbia Basin in Canada and the area draining into the Snake River.
Figure 1. Study Basin in Eastern Washington and Northern Oregon. The basin comprises five, 4-
digit United States Geological Survey (USGS) subbasins that drain into the Columbia River.
3.1 Physiographic Features of the Study Basin
The basin selected contains five, 4-digit and 39, 8-digit USGS basins and extends over
150,404 km2 of territory—about 60% resides in Washington with the rest in Oregon.
In terms of its physiography, the region was developed over basalt materials deposited
millions of years ago.1 Tectonic movements led to the formation of ridges and valleys, while the
eroding force of rivers contributed to wear down of the ridges and redeposit of eroded materials
in valleys, leading to the current physiographic configuration of the region. The topography of the
study basin varies from sandy plains and plateaus to mountain slopes and rocky ridgelines (Figure
2). Elevations range from 150 m to more than 1,000 m above sea level.
The climate of the basin is hot and dry during the summer with maximum temperatures often
exceeding 40°C. Winters bring wet and cold weather with strong winds and blowing snow.
Minimum temperatures in winter often dip to -15°C. The lower Columbia Basin rests deep within
the rain shadow of the Cascade Mountains, so it receives only between 100–230 mm of annual
precipitation (Figure 3), of which about half occurs as snow. Toward the foothills, precipitation
ranges between 400–600 mm.
1http://www.pnl.gov/pals/handbook/part1_1.pdf#search=%22columbia%20basin%20physiography%22.
4
Figure 2. Physiographic Features of the Pacific Northwest. The study basin corresponds
approximately with the Walla Walla plateau.
Figure 3. Geographic Distribution of Annual Precipitation in the Study Region (polygons
delineated in black) and Surrounding Areas. Notice the rain shadow effect east of the Cascade
Mountains (blue and yellow-brown colors along the divortium aquarum of the Cascade
Mountains).
Natural vegetation is typical of desert areas, and it is described broadly as shrub-steppe2.
Dominant shrubs include big sagebrush (Artemisia tridentata Nutt.), spiny hop sage (Grayia
spinosa (Hook.) Moq.), antelope bitterbrush (Purshia tridentata (Pursh) DC.), black greasewood
(Sarcobatus vermiculatus (Hook.) Torr.), and threetip sagebrush (Artemisia tripartita Rydb.).
Large bunchgrasses and flowering forbs make up the rest of the shrub-steppe plant community.
Along the streams, the natural vegetation consists of reeds, rushes, and cattails, as well as
2http://extension.usu.edu/rangeplants/index.htm.
5
deciduous trees and shrubs. The fauna of the region includes about 40 species of mammals, 246
species of birds (some migratory), five species of amphibians, and 10 species of reptiles. The
Columbia River and its tributaries within the study basin serve as habitat for numerous species of
fish (some introduced, others migratory), such as: bluegill sunfish (Lepomis macrochirus), carp
(Cyprinus carpio), channel catfish (Ictalurus punctatus), Chinook salmon (Oncorhynchus
tshawytscha), largemouth bass (Micropterus salmonoides), mosquitofish (Gambusia affinis),
smallmouth bass (Micropterus dolomieui), and rainbow trout (Salmo gairdneri).3
3.2 Population and Economic Activity
Humans have inhabited the Pacific Northwest region for more than 10,000 years. About
3,500 years ago, the inhabitants of the area made a dietary and lifestyle transition from nomadic
communities hunting large animals to sedentary communities relying on salmon fishing for their
sustenance and culture (NRC 2004). Large tribal fisheries existed towards the end of the 18th
century at the Willamette, Cascades, and Celillo Falls in the Columbia River.4 Soon after the
completion of the historic Lewis and Clark expedition, European settlement began early in the
19th century. Since then, the region has undergone significant economic and environmental
transformations with activities such as mining, livestock, dryland agriculture, and timber
harvesting. The construction of numerous dams along the Columbia River for energy production
and irrigated agriculture during the 20th century brought significant economic progress to the
region, but had the adverse effect of altering the normal course of the salmon runs.
Total population in the study basin reached 1 million by 1995, with approximately three
quarters of the population living in Washington and the rest in Oregon.5 The largest urban
development is the Tri-Cities area in the state of Washington, a conglomerate of three cities,
Richland, Kennewick, and Pasco, with an approximate population of 125,000 in 2000.6 The
population of Benton and Franklin counties, where these urban centers are located, totaled about
170,000 people in 1995.7
3.3 Agriculture
Data from the 2000 census reveal that of the total 15,040,400 hectares (ha) representing the
study basin, 1,429,099 ha are used to plant crops, and, of these, 91,864 ha are under irrigation.
However, the U.S. Department of Agriculture (USDA) reported that 111,598 ha bearing orchards
in 1997 were 99% irrigated (Table 1). Data from 2002 show an increase in area for almost all
crops. The USGS reported 895,351 ha of irrigated land in the study basin (1995 USGS
Hydrological Unit Code [HUC] data), which falls well within the range reported by the USDA for
total hectares planted with crops.
In 2000, the USDA did not report if certain commodities (corn, peas, hay, oats, potatoes, and
sugarbeets) within the study basin were irrigated (Table 2).
3http://www.pnl.gov/ecology/Rivers.html. 4http://oregonstate.edu/instruct/anth481/sal/crintro1.htm. 5http://water.usgs.gov/watuse/spread95.html. 6http://en.wikipedia.org/wiki/Tri-Cities,_Washington. 7http://www.workforceexplorer.com/admin/uploadedPublications/385_tricity.pdf.
6
Table 1. 1997 and 2002 USDA Census Data of Irrigated Crops in the Study Basin
Crop No. of Farms (1997) Area (ha) 1997 2002
Apples (1997) 3,265 59,108 71,153
Apricots 256 355 493
Sweet cherries 2,117 13,230 18,310
Cherries (tart) 58 47 437
Grapes 928 21,698 25,332
Hazelnuts (Filberts) 14 2 7
Kiwi fruit 4 0 0
Nectarines 162 409 601
Other fruits and nuts 35 21 40
Peaches, 321 930 1,325
Pears 1,753 15,432 17,371
Plums and prunes 160 354 448
Walnuts 51 13 32
Total 9,124 111,598 135,547
There are other commodities reported as 100% irrigated (all beans) (Table 3), and certain
commodities for which irrigation reporting seemed to be county-dependent (e.g., barley and
wheat).
For barley, irrigation increases yield by 74%. For ―all wheat,‖ the increase is 80%. For spring
wheat, the yield is 170%, and winter wheat yield increases 68% (Table 4). As expected, the yield
responses to irrigation vary by county with Yakima consistently showing the greatest increase in
yield (data not shown).
Table 2. Area Harvested and Production of Various Dryland Agricultural Commodities
Commodity Area harvested (ha) Production (Mg)
Corn For Grain Total 34,034 379,621
Corn For Silage Total 8,296 592,500
Green Peas For Processing Total 14,038 83,600
Hay Alfalfa (Dry) Total 210,841 2,704,300
Hay All (Dry) Total 311,203 3,443,000
Hay Other (Dry) Total 100,362 738,700
Oats Total 2,711 7,750
Potatoes All Total 78,873 5,478,348
Sugarbeets Total 11,331 822,900
7
Table 3. Area Harvested and Production of Various Irrigated Agricultural Commodities
Commodity
Practice
Area harvested
(ha)
Production
(Mg)
Beans Dry Edible All Irrigated equals Total 8,377 21,772
Pink Beans Irrigated equals Total 607 2,087
Pinto Beans Irrigated equals Total 4,047 10,478
Small Red Beans Irrigated equals Total 405 1,089
Small White Beans Irrigated equals Total 283 680
Table 4. Area Harvested and Production of Irrigated and Non-irrigated Agricultural Commodities
Commodity Practice Area harvested (ha) Production (Mg)
Barley Irrigated Total 2,995 13,821
Non Irrigated Total 44,313 117,761
Total For Crop 117,197 381,890
Wheat (all) Irrigated Total 106,270 703,514
Non Irrigated Total 570,526 2,097,436
Total For Crop 1,006,453 4,089,740
Wheat Other Spring Irrigated Total 38,081 250,988
Non Irrigated Total 102,628 250,610
Total For Crop 227,393 741,148
Wheat Winter (all) Irrigated Total 68,190 452,526
Non Irrigated Total 472,309 1,864,614
Total For Crop 779,060 3,348,592
3.4 Water Resources
The Columbia River Basin covers an area of 673,397 km2 from its headwaters in British
Columbia, Canada, to its mouth at Astoria, Oregon. We selected a 150,404 km2 region of the
basin in eastern Oregon and Washington, composing the mainstem of the Columbia River and the
area beginning with the reservoir upstream from the Grand Coulee Dam and ending at the
Bonneville Dam (Figure 1). The average annual flow for the Columbia River at The Dalles,
Oregon, is approximately 5,448 m3 s
-1. The river’s annual discharge rate fluctuates with
precipitation and ranges from 3,171 m3 s
-1 in a low (drought) water year (e.g., 2001) to 7,589 m
3
s-1
in a high water year (e.g., 1997). Land cover changes, particularly the reduced maturity of
forested areas, have altered the hydrology of the river system over the past century, increasing
runoff and reducing evapotranspiration (Matheussen et al. 2000).
The study area has a winter precipitation pattern with two thirds falling between October and
March. Total annual precipitation ranges from 200–600 mm and is strongly dependent on
elevation change (see Figure 2). Historically, multi-year droughts are a typical fluctuation in the
Columbia Basin, with droughts in the 1840s and 1930s ranked as most severe (Gedalof et al.
8
2004). The period of 1950–1987 was unique because of the lack of multi-year drought events.
Further, streamflow in the Columbia River has been linked to large-scale climate fluctuations,
such as the interannual El Niño Southern Oscillation (ENSO) and the Pacific Decadal Oscillation.
Under the warm phase of both events, the region is warmer and drier with lower winter snowfall
and streamflow. During the cold phase, the region is cooler and wetter. Because the area is dry
and dominated by winter precipitation, the primary supply of water is snowpack in the mountain
ranges to the east and west of the central Columbia Basin. This natural reservoir holds the winter
precipitation and releases it throughout spring and into summer. The timing of this snowmelt is
critical to both human activities and salmon survival. Artificial reservoirs have been created
behind dams along almost the entire mainstem of the Columbia River.
Since construction of dams for flood control and power production began in the 1930s, the
flow regime of the river has changed. Records kept since 1878 show that flows were much higher
in the spring and lower in winter, and water velocity was much greater before dam construction.
In 1917, the state of Washington adopted a water code to help manage water allocations. Since
then, it has allocated hundreds of surface and ground water rights on the Columbia River. Water
users have the right to take approximately 1,209 m3 s
-1 in instantaneous withdrawals from April
through October, the growing season for most crops in the basin. The total annual withdrawal
from the mainstem Columbia River during the growing season is about 580,137 m3 of water. The
Bureau of Reclamation is the single largest water user on the river and is allocated about two
thirds of the water.
3.5 A Simple Water Balance of the Study Basin
Based on observed and simulated data from various sources, a simple annual water balance
equation was constructed (Table 5). Overall, there is a good agreement between observed
(Dobserved = 17.2 x 1010
m3 y
-1) and estimated (Destimated = 17.1 x 10
10 m
3 y
-1) streamflows of the
Columbia River at The Dalles, Oregon. Methodologically, this is quite important in developing a
mass balance system to calculate water transactions based on altered streamflows, precipitation,
evapotranspiration, water withdrawals, and in-stream water uses.
3.6 Seasonal Variations of Streamflow and Spatial Distribution of Water Use
Monthly flows and the possible seasonal shifts from potential climate change are more
important than yearly totals given the multi-use aspects of the water from the Columbia River,
i.e., electricity demand from hydropower, irrigation needs during dry periods, and sufficient flows
during the migration of salmon smolts to the ocean.
9
Table 5. Simple Annual Water Balance of the Study Basin Based on Observed Water Flows,
Observed Precipitation, Simulated Evapotranspiration, and Estimated
Water Withdrawals
Water balance
term†
Sources
Data type
Annual flow
(1010
m3 y
-1)
I
Streamflow of Columbia River at the
International Canada-U.S. Border Observed
8.9
S
Streamflow of Snake River
(near Richland, Washington) Observed
4.8
P Precipitation (from the HUMUS model) Observed
7.5
E
Evapotranspiration
(from the HUMUS model) Simulated
3.1
W Water withdrawals (from USGS data) Estimated
1.0
D
Streamflow of Columbia River at The
Dalles, Oregon Observed
17.2
†Explanation of water balance terms: I=streamflow at international border, S=streamflow of Snake River,
P=precipitation, E=evapotranspiration, W=water withdrawals, and D=streamflow at The Dalles. The water
balance equation is D = I + S + P – E – W. Destimated = 17.1 x 1010
m3 y
-1; Dobserved = 17.2 x 10
10 m
3 y
-1.
Table 6 shows monthly streamflow data of the Columbia River at the John Day Dam (i.e.,
mean, high, and low flows) in comparison to monthly water withdrawals (NRC 2004). The NRC
data also give the monthly summed upstream withdrawals at the John Day Dam. In average and
above-average flow years, percentages indicate most water is withdrawn in August. In dry years,
the most water is withdrawn in the spring through July. It is clear from the John Day information
that much of the natural variability of the streamflow is retained, but a maximum of 16.6% of that
flow is withdrawn in July in a dry year in addition to reducing the streamflow by nearly 10% each
of the three months preceding July for agricultural irrigation.
The USGS provides two data sets regarding water supply and demand. One set is based on
county delineations, while the other is based on 4-digit watersheds, or hydrologic unit area 4
(HUA4). Counties and watersheds do not necessarily overlap. However, for the study basin
chosen in this study, the major aspects of the water balance (i.e., irrigation) do not differ
significantly.
Following up on the USGS approach to an irrigation water balance, where return flow is
calculated as total withdrawal for irrigation minus the sum of consumptive use by irrigation and
conveyance losses (http://water.usgs.gov/watuse/tables/irtab.huc.html), we find 10,063 million m3
withdrawn for irrigation, per the 1995 USGS HUA data, and 10,319 million m3 according to the
1995 USGS county data. Those totals include 4,442, million m3 versus 4,535 million m
3 for
consumptive irrigation use, and 1,776 million m3 versus 1,832 million m
3 for conveyance losses,
resulting in 3,845 million m3 versus 3,952 million m
3 left for return flows.
10
Table 6. Monthly and Annual Streamflows of the Columbia River at the John Day Dam and
Monthly and Annual Water Withdrawals from the Columbia River (adapted from
Managing the Columbia River: Instream Flows, Water Withdrawals, and Salmon
Survival, Table 3.1 after unit conversion (NRC 2004))
Mean flow
m3 mo
-1
Maximum
flow
m3 mo
-1
Minimum
flow
m3 mo
-1
Withdrawals
m3 mo
-1
% of
Mean
flow
% of
Max
flow
% of Min
flow
Jan 11,952 19,982 6,698 13 0.10 0.10 0.20
Feb 11,718 22,449 7,080 12 0.10 0.10 0.20
March 13,692 25,163 7,648 136 1.00 0.50 1.80
April 14,925 24,423 7,302 736 4.90 3.00 10.10
May 21,216 36,264 10,004 944 4.40 2.60 9.40
June 23,436 42,802 8,782 977 4.20 2.30 11.10
July 15,419 26,397 6,303 1,048 6.80 4.00 16.60
Aug 10,349 16,529 6,685 978 9.50 5.90 14.60
Sept 7,919 11,422 5,279 614 7.80 5.40 11.60
Oct 8,523 12,828 6,698 338 4.00 2.60 5.00
Nov 9,054 11,447 6,377 15 0.20 0.10 0.20
Dec 10,941 18,626 6,426 15 0.10 0.10 0.20
Annual 159,144
m3 yr
-1
268,332
m3 yr
-1
85,283
m3 yr
-1
5,827
m3 yr
-1
3.70% 2.20% 6.80%
From upstream to downstream, those same water balance calculations are shown as pie charts
for each 8-digit watershed in the left panel of Figure 4. The right panel shows the additional
consumptive-use allocations (commercial, domestic, industrial, mining, livestock, and
thermoelectric) for each of those watersheds.
Of note, the John Day and Deschutes watersheds in Oregon show close to zero industrial
water use, and the reported water use in the John Day watershed is mainly for livestock. Only
electricity generation outside of hydropower occurs in the Middle Columbia, and domestic water
use tends to be more than double commercial water use in these watersheds.
11
Upper Columbia
2383, 49%
1845, 37%
707, 14%
Assumed return flow
Irrigation (consumptive use)
Conveyance losses
Upper Columbia
3.58
8.19
18.210.048.51
0.12
0.00
0.00
13.48
Commercial Water Use
Domestic Water Use
Industrial Water Use
Mining Water Use
Livestock Water Use (Total)
Thermoelectric once through
Thermoelectric closed loop
Thermoelectric geothermal
Thermoelectric nuclear
Yakima
1370, 45%
1129, 38%
498, 17%
Assumed return flow
Irrigation (consumptive use)
Conveyance losses
Yakima
5.82
31.42
0.04
7.21
0.00
0.00
0.00
0.002.25
Commercial Water Use
Domestic Water Use
Industrial Water Use
Mining Water Use
Livestock Water Use (Total)
Thermoelectric once through
Thermoelectric closed loop
Thermoelectric geothermal
Thermoelectric nuclear
Middle Columbia
946, 43%
921, 42%
337, 15%
Assumed return flow
Irrigation (consumptive use)
Conveyance losses
Middle Columbia
7.43
9.160.03
5.42
10.11
0.000.00
0.001.91
Commercial Water Use
Domestic Water Use
Industrial Water Use
Mining Water Use
Livestock Water Use (Total)
Thermoelectric once through
Thermoelectric closed loop
Thermoelectric geothermal
Thermoelectric nuclear
Deschutes
390, 42%
384, 41%
158, 17%
Assumed return flow
Irrigation (consumptive use)
Conveyance losses
Deschutes
6.82
0.00
0.03
2.78
0.00
0.00
0.15
0.000.00 Commercial Water Use
Domestic Water Use
Industrial Water Use
Mining Water Use
Livestock Water Use (Total)
Thermoelectric once through
Thermoelectric closed loop
Thermoelectric geothermal
Thermoelectric nuclear
John Day
103, 30%
163, 48%
77, 22%
Assumed return flow
Irrigation (consumptive use)
Conveyance losses
John Day
0
1
0
0
2
000 0
Commercial Water Use
Domestic Water Use
Industrial Water Use
Mining Water Use
Livestock Water Use (Total)
Thermoelectric once through
Thermoelectric closed loop
Thermoelectric geothermal
Thermoelectric nuclear
Figure 4. Water Balance Calculations (left panel) and Additional Consumptive-use Allocations
(commercial, domestic, industrial, mining, livestock, and thermoelectric) for Watersheds of the
Columbia Basin
12
3.7 Electricity Generation
From an economic modeling perspective, the Pacific Northwest can be considered one
electricity market, with demand for electricity driven by a growing population. Historically,
electricity from hydroelectric dams has been sold to consumers in the Pacific Northwest at prices
less than those from other generating sources. This has resulted in a greater use of electricity for
space heating and a concentration of electricity-intensive industry, such as aluminum smelting.
While aluminum plants demand a steady amount of electricity over hours and months, the shape
of the hourly load profile of the Pacific Northwest electricity system is dominated by space
heating, and the Pacific Northwest is a winter-peaking system.
For planning purposes, the Pacific Northwest hydro system is assumed to supply about
12,000 average Megawatts (aMW). This corresponds to hydroelectric generation available in a
―critical water year‖ or under worst historical conditions. The hydro system supplies about 16,000
aMW in an average water year. Hydro generation of 12,000 aMW is about half of the electricity
generated in the Pacific Northwest. The remainder is generated primarily from natural gas and
coal, with smaller amounts from nuclear and wind (NPCC 2005).
Along with consumptive use of water for irrigation and other minor consumptive-use
demands, water in the study basin is used for hydropower generation. From upstream to
downstream, Table 7 shows the locations and pool heights of the major dams in the study basin,
the hydraulic capacity (106 m
3 y
-1), and the generation capacity (MW) of the dams.
According to the 1995 USGS county data (summed over the study basin), actual electricity
generation amounted to 90,302 GWh, or 91% of the total by hydropower, with an additional
6,942 GW.h by nuclear and 1,731 GW
.h by coal.
As of February 4, 2010, the installed wind capacity in Washington State is 1848.88 MW
(DOE 2010a). In the Middle Columbia, installed wind farms are of 180.2 and 39.6 MW, and 230
MW is slated to come online. As of February 4, 2010, installed wind capacity in Oregon is
1758.14 MW (DOE 2010a). In John Day, wind farms of 24.6, 24, 83.16, and 25.2 MW,
respectively, are installed, totaling 156.96 MW. As of December 31, 2009, the U.S. had 34,863
MW of total installed wind capacity (DOE 2010b).
13
Table 7. Locations and Characteristics of Dams Located on the Columbia River Within the
Study Region
Dam Watershed
Columbia
River
(km)
Spillway
(m)
Pool
height
(High)
Pool
height
(Low)
Hydraulic
Capacity—
First Pump
House
(106 m
3 y
-1)
Name plate
Capacity—
First Pump
House
(MW)
Second
Pump
Grand
Coulee
Upper
Columbia 960 393 368 250,211 6465
Chief
Joseph 877 291 283 195,701 2069
Wells 830 238 235 196,595 774
Rocky
Reach 762 215 214 196,595 1347
Rock
Island 730 187 186 212 410
Wanapum 669 174 171 1038
Priest
Rapids 639 148 147 956
Ice
Harbor
(on Snake
River) Snake 16 180 134
133–
134 94,723 603
McNary
Grande
Ronde;
Yakima 470 399 109
102–
104 207,318 980
John Day John Day 347 374 82 78 287,743 2160
According to The Fifth Northwest Electric Power and Conservation Plan (NPCC 2005),
whole region electricity generation was composed of 52% hydro, 21% natural gas, 20% coal, 3%
biomass, 1% wind, 3% nuclear, and 0% oil. The Fifth Northwest Electric Power and
Conservation Plan (NPCC 2005) also includes the following:
For hydropower most economically and environmentally feasible sites have been developed.
The remaining opportunities are numerous but small scale. The fifth plan calls on utilities to
acquire renewable energy projects including hydropower upgrades as cost-effective opportunities
arise.
14
For the whole Pacific Northwest region, the Bonneville Power Administration (BPA) markets
roughly half of the electricity used in Oregon, Washington, Idaho, and western Montana8,9. The
BPA is a federal agency based in Portland that sells power from 31 federal dams; the Columbia
Generating Station, a non-federal nuclear plant located on the Hanford site in eastern
Washington; and other nonfederal hydroelectric and wind energy generation facilities10
.
Moreover, it controls approximately 75% of the transmission lines in the region.
The BPA is part of the Federal Columbia River Power System (FCRPS), a unique
collaboration among three U.S. government agencies—the BPA, the U.S. Army Corps of
Engineers, and the Bureau of Reclamation. In the past, the BPA has sold power at the cost of
generation with no markup, which is one of the reasons why the Pacific Northwest has enjoyed
the cheapest power rates in the country. This source of inexpensive electricity was a major
attraction for energy-intensive industries, such as aluminum, food processing, and plutonium
production for national defense. In addition, the mining industry was a major beneficiary because
inexpensive electricity greatly reduced the costs of extracting various metals. This resulted in
other industries, such as aerospace, being attracted to the area because they wanted proximity to a
resource (in this case, aluminum) being manufactured in the Northwest.11
However, the aluminum
processing and aerospace industries are located just outside the Columbia River Basin proper, and
only three aluminum plants can be found in the subbasin—a 229 Mton per year plant with an
electricity demand of 428 MW in Chelan County in the Upper Columbia area (Washington); a
166 Mton per year plant with an electricity demand of 317 MW in Klickitat in the Middle
Columbia (Washington); and an 84 Mton per year plant with an electricity demand of 167 MW in
the Deschutes River (Wasco, Oregon).12
Wholesale spot market electricity prices at the Middle Columbia pricing point from January–
December 2003 hovered around $40/MWh (low ~$28; high ~$50) (NPCC 2005). Levelized
annual average electricity price at the Middle Columbia trading hub for 2005 through 2025 is
forecast to be $36.30 per MWh ($2000):
Prices decline between 2005 and 2010 reflecting declining natural gas prices. Prices
increase gradually through the remainder of the planning period as slowly increasing natural gas
prices are partially offset by improved combined-cycle efficiency and increasingly more cost-
effective wind power (NPCC 2005, Vol. 2, p. 25).
3.8 Major Issues—Water, Energy, and Salmon
Since decisions about water typically are not made within a market structure, rather through
administrative or legal frameworks, we have to construct a system that will allow us to
understand how decisions made under a market system might vary from current decisions.
8http://en.wikipedia.org/wiki/Bonneville_Power_Administration. 9http://www.bpa.gov/corporate/. 10www.bpa.gov/corporate/pubs/Keeping/99kc/kc0799.pdf. 11www.fwee.org/c-basin.html. 12The aluminum companies Kaiser, Goldendale Northwest, and Columbia Falls have profited the most from the resale
of BPA power.
15
Typically, water is an input in most of its uses, not a final consumption item. Water for final
consumption actually is a small component of the overall water budget. Therefore, we can make
most of the decisions about water based on the implications of making changes in input levels for
levels of output. The three main uses for water in the Columbia Basin are provision of streamflow
to support salmon, hydroelectric generation, and irrigation. The outputs of water for electricity
can be directly valued using market prices, and we have a good understanding of the productivity
effects for changes in water inputs for these processes to make reliable valuations of the impacts
of changing the available water for these activities. For salmon, we do not have a good sense of
the impact of additional water for streamflow. First, the impact of additional water on the size of
a subsequent returning salmon cohort is highly uncertain. Second, because of the importance of
the non-market value attached to the existence of the salmon fisheries (for which the value of
additional salmon is unclear), the value of an increased cohort size is not well defined. Thus, (Mg
ha-1
) the value of additional salmon reflects both its market value and whatever implications the
fish has for the survival of the fishery. There are methods to define existence value, which are
rough compared to the price of crops and electricity. However, there is no clear idea of how more
fish today affect the future size or existence of the fishery.
The Columbia River Basin has significant variation in annual total flow, as well as a major
variation in flow during the year due to the importance of snowpack as a storage mechanism.
Climate change is likely to reduce snowpack substantially, reducing the ability to manage the
system. Currently, the system can normally be managed to meet the three main users and a
number of other high-value consumptive users, such as residential or manufacturing. It is an open
question if this will be the case under foreseeable increases in regional temperatures due to global
warming.
3.8.1 Electricity
There is more generation capacity on the Columbia River than there is water to power it.
Water removed from the system for irrigation purposes or other consumptive uses reduces the
ability of the system to generate electricity. Because there is a substantial market for peak load
capacity for the California electricity market, the value of the electricity is complex to assess.
Limitations are also imposed by the need to maintain streamflow and temperature to support the
various salmon fisheries. Therefore, computing the value of water for hydroelectric generation
depends on the timing of the reductions.
3.8.2 Irrigation
Irrigation needs are negatively correlated with streamflow, being highest in the period of
lowest flow. Further, different crops respond differently to reductions in available water. For
some crops where the underlying systems require substantial time to reach maturity, there is the
potential for a substantial loss of capital in addition to the loss of current period production.
3.8.3 Salmon
A major problem with salmon is, despite extensive study, the factors that determine salmon
prevalence are not discernable because a major portion of their life cycle takes place in the ocean
16
where factors determining survival are not well understood. Further, beyond the issue of salmon
numbers and the commercial value of the fishery and associated activities, there are existence
values reflected in legislation and regulations mandating certain flows be maintained to preserve
the fisheries.
3.8.4 Decision problem
The question that we wish to pose is how different the water allocation would be if we could
balance the marginal product of water in its different uses. For the purposes of this experiment,
we will treat municipal and industrial uses of water as given (or of such high value that they will
be met). This allows the focus to remain on the major tradeoffs in the basin—irrigation,
hydropower, and the salmon fishery. These tradeoffs occur in a river basin characterized by
highly uncertain flows, which can only be forecast imperfectly. Further, the value of maintaining
streamflow for salmon is also highly uncertain, both in the short run due a poor understanding of
the salmon life cycle in the ocean and in the long term as current river temperatures are already
close to the expected maximum sustainable level.
Given the uncertainty about the benefit of returning more (or less) streamflow for salmon, it
is not possible to balance the return to salmon use against the shadow price of water for irrigation
or hydropower. We can construct two simple experiments to help us understand the extent to
which the current system may have deviated from an optimal economic allocation of water. In the
first experiment, we hold the current allocation of water to maintain the salmon fishery fixed and
analyze the extent (if any) of the imbalance between irrigation and hydropower for three
scenarios—an average year, a high-flow year, and a low-flow year. For high- and low-flow years,
we use the 90 and 10 percentage points on the historical distribution of flows. In the second
experiment, we reallocate all of the water currently reserved for streamflow to irrigation and
hydropower in an optimal format and estimate what the increase in the value of these two streams
would be under the same three scenarios used in the first experiment. This provides an estimate of
the shadow price of the current legal and administrative mechanisms used to protect the salmon
fishery.
In an ancillary analysis, we also examine how climate change may affect the streamflow,
either annually or by changing the within-year distribution of streamflows. This will allow us to
reach a qualitative judgment of the climate’s impact on the shadow prices estimated with the
other two experiments.
3.9 The Future—Climate Change, Legal Issues, and Demographic Changes
3.9.1 Climate change
The water balance of the Columbia River will be impacted in many ways by a changing
climate. Most directly, a change in precipitation amount or timing would considerably alter the
balance of water use for irrigation, energy, and other uses. Indirectly, changes such as an increase
in temperature might drive up the electricity demand of the region, putting increased pressure on
the hydropower system. An analysis of 11 Global Climate Model (GCM) runs for the
17
Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment report by Mote et al.
(2005) produced a suite of possible climate futures for the Pacific Northwest. The range of
warming is projected to be 2–5°C by the end of this century, with the highest increases occurring
in summer. Changes in precipitation, ranging from a 2% decrease to an 18% increase, are not
expected to be distinguishable from natural variability until late in the century. Annual patterns of
change will likely result in an increase in winter precipitation and a decline in summer
precipitation. Under the simulations, the general pattern of winter rainfall will continue.
In general, the projection is for a Pacific Northwest with hotter, drier summers and warmer,
wetter winters. This has substantial implications for water demand, which would likely increase
in the hotter summers as additional irrigation is needed to mitigate heat stress on crops and energy
demands increase. Since natural water storage in snowpacks will decline with increasing
temperature, it also highlights the importance of water storage. These results are consistent with
an earlier study by Payne et al. (2004), which found moderate precipitation changes and a climate
change response dominated by temperature changes. Winter snow accumulation was reduced, and
river flow shifted from summer to winter, causing increased competition for reservoir storage.
The Environmental Policy Integrated Climate (EPIC) model was parameterized with data for
three representative farms in the basin and executed with climate changes from the upper, middle,
and lower parts of this range for periods centered around 2020, 2040, and 2090. The EPIC model
simulates agricultural production and irrigation demand. Each representative farm was simulated
with three crops (wheat, corn, and hay) and for a range of irrigation application scenarios (Figures
5 and 6).
0 2 4 60
600
1200
0
1
2
3
4
5
6
7
8
0 2 4 6
0
600
1200
0
1
2
3
4
5
6
7
8
0 2 4 60
600
1200
0
1
2
3
4
5
6
7
8
0 2 4 60
600
1200
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Yie
ld(M
g h
a-1)
0 1 2 3 4 5 60
600
1200
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0 2 4 60
600
1200
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Irri
ga
tio
n
(mm
)
0 2 4 60
600
1200
0
1
2
3
4
5
6
7
8
9
0 2 4 60
600
1200
0
1
2
3
4
5
6
7
8
9
T Change (C)0 2 4 6
0
600
1200
0123
4
56
7
8
9
Farm: 1702 Farm: 1703 Farm: 1707
Corn
Wheat
Hay
Figure 5. Surface Responses of Crop Yield (Mg ha-1
) as Affected by Increases in Global Mean
Temperatures and Atmospheric CO2 Concentrations in the Columbia Basin
18
The simulations were intended to inform the analysis of the potential range of future
irrigation demand and how much the physical water demand of crops will increase. This
information was used to inform the development of the prototype model as detailed in Section 4.
0 1 2 3 4 5 6
0
400
800
1200
0
5
10
15
20
25
30
35
Yield
T Change
Irrigation
0 1 2 3 4 5 60
500
1000
1500
0
5
10
15
20
25
30
35
40
45
Yield
T Change
Irrigation
Grapes Apples
Figure 6. Surface Responses of Grape and Apple Yields (Mg ha-1
) as Affected by Increases in
Global Mean Temperatures and Atmospheric CO2 Concentrations in the Columbia Basin
3.9.2 Demographics
Population projections for the period 2010–2030 show a 25% increase over a current 6.7
million estimated population for Washington13
and a 19% increase over a current estimated
population of 3.8 million for Oregon14
. If we project and apportion these increases, it would mean
the study basin, by 2030, would have a quarter of a million more people than it does today.
3.9.3 Water rights
In the Columbia River Basin, water is allocated based on an established system of water
rights. New water rights can only be granted through a legal process informed by consideration of
all water demands in the basin, particularly concern for endangered salmon species. A recent
report from the National Academy of Sciences (2004) evaluated the river flows with regard to
ensuring salmon survival and provided recommendations on the amount and timing of water
withdrawals from the system. The structure of water rights may also form a central consideration
of the response of agriculture to climate change. In a study of agriculture in the Yakima River,
Scott et al. (2004) concluded that greater institutional flexibility would be needed to make
effective use of climate forecasts and respond to projected changes in climate.
13http://www.ofm.wa.gov/pop/stfc/default.asp. 14http://www.oregon.gov/DAS/OEA/demographic.shtml#Short_Term_State_Forecast.
19
4.0 Description of Prototype Model
4.1 Description of the Water Prototype Model
4.1.1 Objective
The objective is to build a modeling capability that can be applied to understand how water is
allocated within a river basin and develop a system for modeling present and future water
allocations among agriculture, energy production, other human requirements, and ecological
needs.
4.1.2 Modeling platform
The Water Prototype Model (WPM) was built in STELLA®15
, a computer modeling package
with an easy, intuitive interface that allows users to construct dynamic models that realistically
simulate and integrate many processes (biological, hydrological, economics, sociological), as
described by Costanza and Voinov (2001):
STELLA includes a procedural programming language that is useful to view and analyze the
equations that are created as a result of manipulating the icons. The essential features of the
system are defined in terms of stocks (state variables), flows (in and out of the state variables),
auxiliary variables (other algebraic or graphical relationships or fixed parameters), and
information flows. Mathematically, the system is geared towards formulating models as systems
of ordinary differential equations and solving them numerically as difference equations. The user
places the icons for each of the stocks in the modeling area and then connects them by flows of
material or informational relationships. Next the user defines the functional relationships that
correspond to these flows. These relationships can be mathematical, logical, graphical, or
numerical.
4.1.3 Model description
A general diagram of the model is shown in Figure 7, and a display of the results produced by
the model, including a picture of the modeled basins, is shown in Figure 8. As previously
described, the study basin occupies 150,404 km2 and is part of the larger Columbia River Basin in
the U.S. Pacific Northwest region. About 60% of the study basin rests in the state of Washington,
and the rest is in Oregon. The Columbia River runs through the basin for 874 km starting at the
international border with Canada (49° 00′, 117° 38′) and ending (for the purpose of the
simulation) at The Dalles Dam (45° 37′, 121° 08′). Water enters the basin through precipitation
and from streamflows originating from the Columbia River at the international border, the
Spokane River, and the Snake River. Water leaves the basin through evapotranspiration,
consumptive uses (irrigation, livestock, domestic, commercial, mining, industrial, and off-stream
power generation), and streamflow through The Dalles Dam. Water also enters the Columbia
River via runoff from land (both surface and subsurface). The model runs on a monthly time scale
15http://www.iseesystems.com/softwares/Education/StellaSoftware.aspx.
20
to account for the impact of seasonal variations of climate (interannual variability or ENSO and
climate change), streamflows, and water uses. A salmon policy feature was included to capture
the influence of flow augmentation on smolts migration to the ocean and the consequent
reduction on hydropower production.
Figure 7. Model Diagram Showing the Larger Compartments (i.e., land, river, etc.) and the
Stocks and Flow Occurring Throughout
21
Figure 8. Screenshot of the Water Prototype Model Interface
22
4.1.4 Fundamental equations
The fundamental equations used to build the model concern the storage and transfer of water
within the study basin.
WYETPt
B
[1]
where ∂B/∂t = change in basin storage (m3), P = precipitation (m
3), ET = evapotranspiration
(m3), and WY = water yield (m
3).
)( ETPBfWY t [2]
where f = fraction (dimensionless) and Bt = basin storage at time t (m3). The value of f was
derived from the WY output.
Due to lack of a clear difference between climate change scenarios, water yield was made a
function of ENSO scenarios but not of climate change.
The water content of the basin (Bi, m3) was initialized at 50% of the available soil water
capacity to a depth of 2 m. Data for this calculation were derived from the soil database residing
in the hydrological HUMUS model (Thomson et al. 2003 and 2005b).
CUSSt
Soi
[3]
where ∂S/∂t = change in water volume held by dams (m3), ΣSi = total incoming streamflow
from tributaries outside basin (m3), So = streamflow at basin outlet (m
3), and ΣCU = total
consumptive use by different sectors (m3). Data for the monthly values of CU were derived from
USGS databases for the year 1995, the last reporting period with water use data reported at the 8-
digit HUC available at the writing of this report.
The change in water storage ∂S/∂t was assumed to be 0. Thus, So was calculated as:
CUSS io [4]
In case ∂S/∂t ≠ 0, So can be recalculated based on extra water additions or withdrawals.
The monthly variability of So was modeled by imposing ENSO and climate change scenarios
to baseline conditions existing for P, ET, and streamflows. Two ENSO scenarios are represented
in the WPM model, La Niña and El Niño (Thomson et al. 2003). The two scenarios of climate
change were taken from results from the HUMUS model (Thomson et al. 2005b), Australian
Bureau of Meteorological Research Centre (BMRC), and University of Illinois, Urbana
Champaign (UIUC) for a 1°C increase in global mean temperature (GMT).
23
Consumptive use due to irrigation was made a function of ENSO and climate change
scenarios. A scenario of biodiesel production was also included to capture the influence of
bioenergy production in the Pacific Northwest on water demand. The scenario is based on
biodiesel production from irrigated canola. Data used to calculate the water demand include:
canola yield, water use efficiency, oil concentration in canola seed, biodiesel production from
canola oil, and irrigation requirements.
Calculation of hydropower (HP, W) was calculated as:
gEShHP [5]
where h = dam height (m), S = streamflow (m3 s
-1), E = efficiency factor (dimensionless), g =
acceleration of gravity (m s-2
), and δ = density of water (kg m-3
).
To simulate the effect of flow augmentation on hydropower production, a simple procedure
was added to the model. Flow augmentation is a policy option available to facilitate the migration
of salmon smolts from upstream to the ocean. At the moment, the model does not consider the
effect of other features, such as barging. Future updates of the WPM will include a simplified
version of the Smolts Migration Model developed by the BPA.16
This model calculates the travel
time and survival of smolts batches released from hatcheries as affected by high or low flow,
mortality, barging, turbines, and flow augmentation.
4.1.5 Data sources
Precipitation, evapotranspiration, and water yield data were extracted from databases
containing results obtained with the HUMUS model runs for the conterminous U.S. under
baseline, ENSO, and climate change scenarios (Thomson et al. 2003 and 2005b). Consumptive-
use data were derived from the USGS water use database17
for 1995 available at the 8-digit HUC
level and aggregated across the basin.
Hydropower data—dam height (m), nameplate capacity (kW), and hydraulic capacity (m3 s
-
1)—were obtained from various sources, including the U.S. Department of Interior and the U.S.
Army Corps of Engineers. A weighted average approach based on active dam capacity (m3) was
used to calculate the height of a composite dam made of 10 dams: Grand Coulee, Chief Joseph,
Wells, Rocky Reach, Rock Island, Wanapum, Priest Rapids, McNary, John Day, and The Dalles.
The surface areas of the lakes above each dam were calculated using width and length values
estimated with Google Earth.
4.1.6 How to run the model
The WPM can be run from three sources: 1) directly from STELLA software, 2) with the isee
Player®, or 3) the web version of WPM constructed with NetSim
® software posted at
http://forio.com/broadcast/netsim/netsims/rcizaurra/waterprototype/index.html.
16http://www.bpa.gov/Corporate/KR/ed/step/smolts/SmoltsM.shtml. 17http://water.usgs.gov/watuse/.
24
When running any of these three versions, the user is presented a screen with a series of
buttons, graphs, and tables (see Figure 8). Two of the buttons provide the user with background
and instructions on how to run the model. Currently, there are five types of scenarios that can be
manipulated alone or in combination using the Sliding Input Devices: 1) ENSO, 2) climate
change (ClimScen), 3) salmon policy, 4) future population (FuturePop), and 5) biodiesel
production (MGalBiodiesel).
The ENSO scenario allows for three choices: 1) all years, 2) La Niña years, and 3) El Niño
years.
The climate change scenarios are: 0) baseline, 1) predictions from BMRC GCM for 1°C
increase in GMT, and 2) predictions from UIUC GCM for 1-degree increase in GMT.
The salmon policy slide rule allows the user to set a fraction of streamflow away from
hydropower production and redirect it to help smolts travel to the ocean during the spring. The
biodiesel slider input device allows the user to select different levels of biodiesel production,
from zero to 20 million gallons per year. Based on this information, the model calculates the extra
water demand during the growing season.
The future population (FuturePop) input device allows the user to select a population increase
(1.0–1.3) that will affect the consumptive uses by domestic, commercial, industrial, mining, out-
of-stream power, and livestock sectors.
Finally, the biodiesel policy input device (MGalBiodiesel) allows the user to select a level of
annual production of biodiesel (0–20 million gallons of biodiesel per year) from irrigated canola
and its impact on water demand.
4.1.7 Examples and discussion of selected model outputs
One of the fundamental premises for this capability development activity was to build a
model prototype capable of describing the supply and demand of water at the basin scale yet to be
simple and scalable enough for its inclusion into an integrated model, such as MiniCAM. Figure
9 compares predicted and observed streamflow at The Dalles under average weather conditions,
as well as those of La Niña and El Niño. In general, the predicted monthly flows agree quite well
with the observed values. This was achieved through the integration of observations with
modeled data and predictions of hydrological variables. The model also responded to scenarios of
climate change (data not shown). Although due to the climate change scenarios selected, the
changes in streamflow were not dramatically different from those of the baseline conditions.
One of the important dynamics from the perspective of IA modeling is the consequence of
changes in the prototype variables for hydropower production. Under La Niña conditions, more
hydropower is available during all months of the year, with a substantially higher availability in
the spring and summer months. Conversely, under El Niño conditions, hydropower will be less
available, with a total decline of 15% from normal weather conditions over the year. By contrast,
imposing a policy where water remains in the river for salmon migration has a smaller, but
negative, impact on the total annual supply of hydropower. Modeled hydropower generation was
23% greater than the 81 TWh reported in the 1995 USGS database (Figure 10).
25
Figure 9. Predicted and Observed Streamflows (m3) at The Dalles as Affected by
ENSO Scenarios
Figure 10. Monthly Hydropower (kWh) as Affected by ENSO Scenarios
26
This prototype can be applied to gain understanding of the quantitative dynamics of water
supply and the consequences for different economic sectors, including withdrawals for industry,
energy use, and energy production in the region. In particular, it can be used in the design of an
IA water module to provide insight into the factors affecting allocation of available water among
the competing demands. Once this allocation is understood, economic drivers and controls on
water can be incorporated into the prototype. Then, this will inform the process of applying prices
to water in this and other regions for IA modeling.
4.1.8 Summary and future steps
A WPM to describe water allocation within a basin among multiple uses (agriculture, energy
production, other human requirements, and ecological needs) has been presented and discussed.
The modeling capability is considered the initial step to incorporate water into IA models. The
WPM allows for the analysis of the interactions among water uses in a manner that could be
directly assimilated into IA frameworks. However, to be practical for use in integrated
assessments, it would have to be expanded or scaled up, first nationally, then to the global level.
As demonstrated with the WPM, a spatial scale of a 2-digit basin is considered appropriate to
represent water supply and demand issues for IA. In the case of the conterminous U.S., this would
translate into scaling the WPM to represent 18 major river basins. All of the data and model
results are available to accomplish this model expansion.
However, specific model enhancements would be required to bring realism to the dynamics
of water in different basins. One example is groundwater extractions. Currently, the WPM does
not contain a way to account for water extractions from groundwater resources, yet the
underlying data exist (USGS databases) for representing this process. Another example is water
demand for cellulosic ethanol production, especially involving bioenergy crops such as
switchgrass (Panicum virgatum L.) and miscanthus (Miscanthus x giganteus), as well as water
demand associated with biorefineries.
By adding a one-reservoir submodel to describe the spring migration of smolts to the ocean
and associated policies, the salmon policy currently available in the WPM will be updated in the
near future. The essential features of this submodel are presented in Figure 11. The one-reservoir
submodel is a simplification of the BPA’s more complex model, which includes several
reservoirs along the Snake and Columbia rivers.
As presented, the WPM offers all of the necessary ingredients to conduct economic analyses
of various policy options for water allocation. Even the difficult question of the valuation of water
allocation for salmon survival can be indirectly estimated through its impact on hydropower
production or water withdrawals for irrigation. Further work is needed in this area.
27
Figure 11. Simplified Version of the Bonneville Power Administration Model to Describe Smolts
Migration to the Ocean
28
5.0 References
Costanza R and A Voinov. 2001. ―Modeling ecological and economic systems with
STELLA: Part III.‖ Ecological Modelling 143(1-2):1-7.
DOE – U.S. Department of Energy. 2010a. Installed Wind Capacity by State. Accessed
November 17, 2010 at http://www.windpoweringamerica.gov/wind_installed_capacity.asp (last
updated February 4, 2010).
DOE – U.S. Department of Energy. 2010b. Current Installed Wind Capacity Map. Accessed
November 17, 2010 at
http://www.windpoweringamerica.gov/images/windmaps/installed_capacity_current.jpg (last
updated December 31, 2009).
Gedalof Z, DL Peterson, and NJ Mantua. 2004. ―Columbia River Flow and Drought Since
1750.‖ JAWRA Journal of the American Water Resources Association 40(6):1575-1592.
Izaurralde RC, NJ Rosenberg, RA Brown Jr, and AM Thomson. 2003. ―Integrated
Assessment of Hadley Centre (HadCM2) Climate-Change Impacts on Agricultural Productivity
and Irrigation Water Supply in the Conterminous United States. Part II. Regional Agricultural
Production in 2030 and 2095.‖ Agricultural and Forest Meteorology 117(1-2):97-122.
Matheussen B, RL Kirschbaum, IA Goodman, GM O’Donnell, and DP Lettenmaier. 2000.
―Effects of land cover change on streamflow in the interior Columbia River Basin (USA and
Canada).‖ Hydrological Processes 14(5):867-885.
Mote PM, EP Salathé, and C Peacock. 2005. Scenarios of Future Climate for the Pacific
Northwest. A report prepared for King County Department of Natural Resources by the Climate
Impacts Group, Center for Science in the Earth System, Joint Institute for the Study of the
Atmosphere and Ocean, 13 pp, University of Washington, Seattle.
NPCC – Northwest Power and Conservation Council. 2005. The Fifth Northwest Electric
Power and Conservation Plan. Council Document 2005-07, Portland, Oregon. Accessed
November 12, 2010 at http://www.nwcouncil.org/energy/powerplan/5/Default.htm.
NRC – National Research Council. 2004. Managing the Columbia River: Instream Flows,
Water Withdrawals, and Salmon Survival. National Academies Press, Washington, D.C.
Payne JT, AW Wood, AF Hamlet, RN Palmer, and DP Lettenmaier. 2004. ―Mitigating the
Effects of Climate Change on the Water Resources of the Columbia River Basin.‖ Climatic
Change 62(1-3):233-256.
Rosenberg NJ, RA Brown, RC Izaurralde, and AM Thomson. 2003. ―Integrated Assessment
of Hadley Centre (HadCM2) Climate Change Projections on Agricultural Productivity and
Irrigation Water Supply in the Conterminous United States. I. Climate change scenarios and
impacts on irrigation water supply simulated with the HUMUS model.‖ Agricultural and Forest
Meteorology 117(1):73-96.
29
Sands RD and JA Edmonds. 2005. ―Economic analysis of field crops and land use with
climate change.‖ Climatic Change 69(1):127-150.
Sands, R.D., and M. Leimbach. 2003. ―Modeling Agriculture and Land Use in an Integrated
Assessment Framework,‖ Climatic Change 56 (1): 185-210.
Scott MJ, LW Vail, J Jaksch, CO Stöckle, and A Kemanian. 2004. ―Water Exchanges: Tools
to Beat El Niño Climate Variability in Irrigated Agriculture.‖ JAWRA Journal of the American
Water Resources Association 40(1):15-31.
Thomson AM, RA Brown Jr, NJ Rosenberg, RC Izaurralde, D Legler, and R Srinivasan.
2003. ―Simulated Impacts of El Niño/Southern Oscillation on United States Water Resources.‖
Journal of the American Water Resources Association 39(1):137-148.
Thomson AM, RA Brown, NJ Rosenberg, RC Izaurralde, and V Benson. 2005a. ―Climate
Change Impacts for the Conterminous USA: An Integrated Assessment Part 3: Dryland
Production of Grain and Forage Crops.‖ Climatic Change 69(1):43-65.
Thomson AM, RA Brown, NJ Rosenberg, R Srinivasan, and RC Izaurralde. 2005b. ―Climate
Change Impacts for the Conterminous USA: An Integrated Assessment Part 4: Water Resources.‖
Climatic Change 69(1):67-88.
Thomson AM, NJ Rosenberg, RC Izaurralde, and RA Brown. 2005c. ―Climate Change
Impacts for the Conterminous USA: An Integrated Assessment Part 6: Distribution and
Productivity of Unmanaged Ecosystems.‖ Climatic Change 69(1): 89-105.
